[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Condor-users] retrieving output subdirectories



Mark,

Nice solution you have.  What I would like to do is to make use of GridSite and drop a .gacl file into each execute directory.  The .gacl file could be built dynamically when the job is submitted, possibly with input from the job submitter.  Aliasing or web-proxies could be used to give "nice" URLs to job output directories, or require a shared file system which is web-accessible.

Ian.

Mark Calleja wrote:
Hi Ian,

Ian Stokes-Rees wrote:
  
And a related question: is there any way to view output files for jobs 
that are in the process of executing?
    

Do you mean for the vanilla universe? In that case some of us have 
rolled our own solutions, e.g. this is how we do it:

http://www.escience.cam.ac.uk/projects/camgrid/file-viewer.html

There are other ways too. At last November's HTC workshop in Edinburgh, 
Miron mentioned somebody else's solution that had a process that got 
submitted with a job that could be interrogated on the fly, but I'm not 
aware of any documentation. Miron, have you got any more details of this 
method?

On a broader scope, are there any other solutions out there that people 
would like to mention? I'm very interested in this subject and would 
love to hear of other approaches.

Cheers,
Mark

  

-- 
Ian Stokes-Rees                            W: http://sbgrid.org
ijstokes@xxxxxxxxxxxxxxxxxxx               T: +1 617 418-4168
SBGrid, Harvard Medical School             F: +1 617 432-5600